Hi, I just finished reading this article (https://bytecodealliance.org/articles/wasmtime-and-cranelift-in-2023) about Wasmtime & Cranelift. There was a call for help with respect to a planned interpreter tier for Wasmtime. What is planned for this? What are the requirements for this interpreter tier? What is going to be its purpose and focus? Is there already existing work/branch or a PoC? How can people like me help? Thanks. :)
So far I have worked on wasmi
which is a rewriting Wasm interpreter with a focus on a balance between compile performance and execution performance.
@fitzgen (he/him) [on vacation until Dec. 10] may be the best to answer this, but he's out on vacation currently
Is @fitzgen (he/him) back for an answer? :)
Hey @Robin Freyler, yes I'm back now.
What is planned for this? What are the requirements for this interpreter tier? What is going to be its purpose and focus?
This all needs to be hashed out in an RFC by the stakeholders who intend to drive it forward. I think this might be @Bailey Hayes, but as far as I know, no one is quite ready to dive in head first yet.
How can people like me help?
I think the thing we need now are clarity and agreement on exactly the kinds of things you are probing about from the people who are interested in an interpreter tier and intend to actually implement and maintain it long term, so connecting with them and starting an RFC is probably the best way to push this forward.
Hi @fitzgen (he/him) thanks for your reply!
From what I grapsed there isn't much information, yet. I was asking because I am currently just looking for ways to collaborate more closely with the Bytecode Alliance and was recently made aware of the interpreter tier ideas. My experience is with building Wasm interpreters so I thought I could maybe help out a bit.
Last updated: Nov 22 2024 at 16:03 UTC